流数据决策单元效率评价方法与应用研究

流数据是一种快速抵达并且需要近乎实时处理的数据。现如今，作为大数据最重要来源之一的流数据广泛存在并已成为业界关注的热点。数据包络分析（DEA）作为效率评价的重要工具，已取得极大的理论发展并被广泛应用于各个领域。现实中，流数据决策单元的效率评价问题大量存在，而现有的DEA 效率评价方法无法对其进行处理。因此，本项目拟开展流数据决策单元的效率评价方法研究。根据流数据的特征，主要内容包括：（1）流数据持续更新时的决策单元效率评价方法；（2）流数据低价值密度时的决策单元效率评价方法；（3）流数据动态变化时的决策单元效率评价方法；（4）上述方法在互联网金融行业绩效评价中的应用，重点考查金融机构的经营绩效并应用于客户欺诈识别、客户流失预防和精准营销、客户行为预测分析等。这些研究将进一步推动 DEA 方法的理论发展以及应用实践，并为企业的管理决策提供支持。

Streaming data arrives at an unprecedented speed and must be dealt with in a timely manner. Nowadays, streaming data becomes the focus of the industry and enterprises. It is widespread and is one of the most important sources of big data. Data envelopment analysis (DEA), which is an important tool for efficiency evaluation, has made great progresses in methodology research and is widely used in many areas. In the real world, there are a lot of problems that are related to efficiency evaluation for decision making units (DMUs) with streaming data. However, existing DEA methods for efficiency evaluation cannot handle them. As a result, this project will examine the problem of efficiency evaluation for the DMUs with streaming data. According to the characteristics of streaming data, the project composes of four categories: (1) the efficiency evaluation for the DMUs when streaming data is updated continuously; (2) the efficiency evaluation for the DMUs when the value density of streaming data is low; (3) the efficiency evaluation for the DMUs when streaming data is changed dynamically; (4) the applications of the above methodologies in the efficiency evaluation for the Internet financial industry. The emphasis is to study the business performance of financing institutions and apply the proposed methods to customer fraud identification, customer loss prevention and precise marketing and customer behavior prediction analysis, etc. These research will further promote the theoretical developments and practical applications of DEA, and provide substantial supports for management decision of enterprises.